[关键词]
[摘要]
在模型预测控制中,大量的计算会产生动作延迟,并且离散化方法的不准确性也会将误差引入系统,当采样时间较长时上述两种情况会导致系统性能恶化。为解决该问题,可采用双线性变换法获得更为精确的电机离散时间模型,该离散方法可保证系统有更优的稳定性。再通过引入延时补偿来解决预测电流时由于计算导致的动作延迟,从而减小电流纹波。基于此搭建仿真模型并进行试验研究,仿真与试验证明该预测系统具有动态跟踪特性好、电流纹波小和抗负载扰动强等优点。
[Key word]
[Abstract]
In model predictive control system, the large amount of calculation could lead to action delay and the error would be introduced to system due to the inaccuracy discrete method. Under the condition of large sample time, the two conditions above would lead to a bad performance of the system. To solve the problem, the bilinear transformation method was taken to obtain a more accurate discrete model of motor. And this discrete method could guarantee the stability of system. Delay compensation was introduced to the system to reduce the current ripple due to the problem of action delay. Simulation and experiment results showed that the predictive system above had the advantages of better dynamic track performance, little current ripple and strong resistance to load disturbance.
[中图分类号]
TM 351
[基金项目]